Phase I, Open-Label Study of Pasireotide in Patients with Wild Type and -Wild Type, Unresectable And/Or Metastatic Melanoma

Zurich Open Repository and Archive University of Zurich Main Library Strickhofstrasse 39 CH-8057 Zurich www.zora.uzh.ch

Year: 2018

Phase I, open-label study of pasireotide in patients with wild type and -wild type, unresectable and/or metastatic melanoma

Dummer, Reinhard ; Michielin, Olivier ; Nägeli, Mirjam Chantal ; M Goldinger, Simone ; Campigotto, Federico ; Kriemler-Krahn, Ulrike ; Schmid, Herbert ; Pedroncelli, Alberto ; Micaletto, Sara ; Schadendorf, Dirk

Abstract: Introduction Somatostatin analogues exert antitumour activity via direct and indirect mechanisms. The present study was designed to assess the safety and eﬀicacy of pasireotide in patients with -wild type (WT) and -WT metastatic melanoma. Patients and methods Patients with unresectable and/or metastatic melanoma or Merkel cell carcinoma were eligible. Pasireotide was administered at different doses for ฀8 weeks in dose-escalation phase, followed by long-acting pasireotide 80 mgorlower dose in case of toxicity in follow-up phase up to six additional months. Primary endpoint was safety in the first 8 weeks of dose-escalation phase. Results The study was terminated early due to slow recruitment. Of the 10 patients with metastatic melanoma enrolled, only four reached the high dose level: two patients reached 3600 µg in dose-escalation and follow-up phases and two patients reached 3600 µg in dose-escalation and long-acting pasireotide 80 mg in follow-up phases and were stable for >5 months. Most common adverse events (AEs) during dose-escalation phase in ฀2 patients (20%) were: diarrhoea (50%), nausea (50%), fatigue (20%), hyperglycaemia (20%), hypophosphatemia (20%), chills (20%) and tumour pain (20%). Grade 3 or 4 study drug-related AEs were diarrhoea and nausea, reported in one patient. Partial response was documented in one patient and stable disease in another. Conclusions Pasireotide was well tolerated, and safety results were similar to those previously reported in other indications. Further studies are needed to evaluate its antitumour activity alone and in combination with other drugs in melanoma.

DOI: https://doi.org/10.1136/esmoopen-2018-000388

Posted at the Zurich Open Repository and Archive, University of Zurich ZORA URL: https://doi.org/10.5167/uzh-153390 Journal Article Published Version

The following work is licensed under a Creative Commons: Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License.

2 Zurich Open Repository and Archive University of Zurich Main Library Strickhofstrasse 39 CH-8057 Zurich www.zora.uzh.ch

Year: 2018

Phase I, open-label study of pasireotide in patients with wild type and -wild type, unresectable and/or metastatic melanoma

Dummer, Reinhard; Michielin, Olivier; Nägeli, Mirjam Chantal; M Goldinger, Simone; Campigotto, Federico; Kriemler-Krahn, Ulrike; Schmid, Herbert; Pedroncelli, Alberto; Micaletto, Sara; Schadendorf, Dirk

Abstract: Introduction Somatostatin analogues exert antitumour activity via direct and indirect mechanisms. The present study was designed to assess the safety and efficacy of pasireotide in patients with -wild type (WT) and -WT metastatic melanoma. Patients and methods Patients with unresectable and/or metastatic melanoma or Merkel cell carcinoma were eligible. Pasireotide was administered at different doses for ฀8 weeks in dose-escalation phase, followed by long-acting pasireotide 80 mgorlower dose in case of toxicity in follow-up phase up to six additional months. Primary endpoint was safety in the first 8 weeks of dose-escalation phase. Results The study was terminated early due to slow recruitment. Of the 10 patients with metastatic melanoma enrolled, only four reached the high dose level: two patients reached 3600 µg in dose-escalation and follow-up phases and two patients reached 3600 µg in dose-escalation and long-acting pasireotide 80 mg in follow-up phases and were stable for >5 months. Most common adverse events (AEs) during dose-escalation phase in ฀2 patients (20%) were: diarrhoea (50%), nausea (50%), fatigue (20%), hyperglycaemia (20%), hypophosphatemia (20%), chills (20%) and tumour pain (20%). Grade 3 or 4 study drug-related AEs were diarrhoea and nausea, reported in one patient. Partial response was documented in one patient and stable disease in another. Conclusions Pasireotide was well tolerated, and safety results were similar to those previously reported in other indications. Further studies are needed to evaluate its antitumour activity alone and in combination with other drugs in melanoma.

DOI: https://doi.org/10.1136/esmoopen-2018-000388

Posted at the Zurich Open Repository and Archive, University of Zurich ZORA URL: https://doi.org/10.5167/uzh-153390 Journal Article

Originally published at: Dummer, Reinhard; Michielin, Olivier; Nägeli, Mirjam Chantal; M Goldinger, Simone; Campigotto, Fed- erico; Kriemler-Krahn, Ulrike; Schmid, Herbert; Pedroncelli, Alberto; Micaletto, Sara; Schadendorf, Dirk (2018). Phase I, open-label study of pasireotide in patients with wild type and -wild type, unresectable and/or metastatic melanoma. ESMO Open, 3:e000388. DOI: https://doi.org/10.1136/esmoopen-2018-000388 Open access Original research ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from Phase I, open-label study of pasireotide in patients with BRAF-wild type and NRAS-wild type, unresectable and/or metastatic melanoma

Reinhard Dummer,1 Olivier Michielin,2 Mirjam Chantal Nägeli,1 Simone M. Goldinger,1 Federico Campigotto,3 Ulrike Kriemler-Krahn,4 Herbert Schmid,4 Alberto Pedroncelli,4 Sara Micaletto,1 Dirk Schadendorf5

► Additional material is ABSTRACT Key questions published online only. To view Introduction Somatostatin analogues exert antitumour please visit the journal online activity via direct and indirect mechanisms. The present (http://dx. doi. org/ 10. 1136/ What is already known about this subject? study was designed to assess the safety and efficacy esmoopen-2018- 000388). Until now, the treatment options for patients with of pasireotide in patients with BRAF-wild type (WT) and ► BRAF-wild type (WT) and NRAS-WT melanoma are NRAS-WT metastatic melanoma. To cite: Dummer R, Michielin O, limited. Nägeli MC, et al. Phase I, Patients and methods Patients with unresectable and/ ► Mitogen-activated protein kinase (MAPK) and PI3K open-label study of pasireotide or metastatic melanoma or Merkel cell carcinoma were pathways are the major signalling pathways in mel- in patients with BRAF-wild eligible. Pasireotide was administered at different doses anoma, which regulate cell growth, proliferation, type and NRAS-wild type, for ≤8 weeks in dose-escalation phase, followed by long- survival and transformation. Somatostatin recep- unresectable and/or metastatic acting pasireotide 80 mg or lower dose in case of toxicity tors (SSTRs) and insulin growth factor receptors are melanoma. ESMO Open in follow-up phase up to six additional months. Primary overexpressed in human melanoma cells and are 2018;3:e000388. doi:10.1136/ endpoint was safety in the first 8 weeks of dose-escalation esmoopen-2018-000388 upstream of the Ras/MAPK signalling pathway. phase. Somatostatin analogues (such as pasireotide) exert Results The study was terminated early due to slow ► antitumour activity in different tumours like pituitary, recruitment. Of the 10 patients with metastatic melanoma Received 25 April 2018 prostate, gastric, lung, pancreas, colorectal or thy- Revised 25 May 2018 enrolled, only four reached the high dose level: two roid origin, both via direct (mediated through SSTRs) Accepted 26 May 2018 patients reached 3600 µg in dose-escalation and follow-

and indirect mechanisms (insulin-like growth factor http://esmoopen.bmj.com/ up phases and two patients reached 3600 µg in dose- 1 (IGF-1) signalling pathway). © European Society for escalation and long-acting pasireotide 80 mg in follow-up Medical Oncology 2018. Re-use phases and were stable for >5 months. Most common What does this study add? permitted under CC BY-NC. No adverse events (AEs) during dose-escalation phase in ► This is the first phase I study to assess the safety and commercial re-use. Published ≥2 patients (20%) were: diarrhoea (50%), nausea (50%), efficacy of pasireotide in patients with unresectable by BMJ. fatigue (20%), hyperglycaemia (20%), hypophosphatemia and metastatic BRAF-WT and NRAS-WT melanoma. 1 Department of Dermatology, (20%), chills (20%) and tumour pain (20%). Grade 3 or ► We observed that pasireotide was well tolerated, University Hospital Zurich, 4 study drug-related AEs were diarrhoea and nausea, and the safety profile was similar to the prior reports Zurich, Switzerland reported in one patient. Partial response was documented in other indications. 2Multidisciplinary Oncology in one patient and stable disease in another. ► Of 10 patients included in the study, stable disease on 29 August 2018 by guest. Protected copyright. Center, Lausanne University Conclusions Pasireotide was well tolerated, and safety Hospital (CHUV), Lausanne, and partial response were observed in one patient results were similar to those previously reported in other Switzerland each. 3 indications. Further studies are needed to evaluate its Global Medical Affairs, Novartis How might this impact on clinical practice? Pharmaceuticals Corporation, antitumour activity alone and in combination with other East Hanover, New Jersey, USA drugs in melanoma. ► Though the antitumour activity in these late-staged 4Clinical Development, Novartis patients was limited, pasireotide could be potentially Pharma AG, Basel, Switzerland used in combination with other therapeutic agents in 5Department of Dermatology, IntRoduCtIon the treatment of patients with BRAF-WT and NRAS- University Hospital Essen, Melanoma is the most aggressive and some- WT melanoma. Essen, Germany & German times treatment-resistant form of skin cancer. ► Further prospective randomised studies are war- Cancer Consortium, Heidelberg, ranted to evaluate the efficacy of pasireotide in BRAF-wild type (WT) and NRAS-WT tumours Germany BRAF-WT and NRAS-WT melanoma. are observed in approximately 28%–32% Correspondence to of melanomas.1 2 The constitutive activa- Professor Reinhard tion of the mitogen-activated protein kinase Dummer, Department of BRAF DermatologyUniversity Hospital (MAPK) signalling pathway in -WT and (a tumour suppressor gene encoding the RAS Zurich Zurich Switzerland ; NRAS-WT melanoma may be due to the loss GTPase-activating protein). NF1 mutations reinhard.dummer@ usz. ch of function mutations and deletions in NF1 are identified in up to 70% of BRAF-WT and

Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 1 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from

NRAS-WT melanomas.3 The other activating mutations in Solid Tumour version 1 (RECIST 1.0) and the pres- of KIT, amplification of KIT, CCND1 and TERT are also ence of ≥1 lesion suitable for standardised uptake value observed in this population.4 5 measurements on 18-fluorodeoxyglucose positron emis- Somatostatin and its analogues (SSAs) bind to soma- sion tomography (18-FDG-PET) were included in the tostatin receptors (SSTR)1–5 with variable affinities study. The other key eligibility criteria included Eastern and may directly inhibit the cell growth by the inhibi- Cooperative Oncology Group performance status of 0 or tion of cell proliferation through the cell cycle control 1 and adequate organ function. Patients with unknown and induction of apoptosis.6 The data from preclinical BRAF-mutation or NRAS-mutation status, primary uveal and clinical studies demonstrate that SSAs exert antitu- melanoma, prior treatment with SSA or hypersensitivity mour effects in different tumour types including those to SSA and those receiving >3 prior lines of systemic of pituitary, prostate, gastric, lung, pancreas, colorectal therapy, prior radiotherapy, an investigational drug or or thyroid origin.7–11 One or more subtypes of SSTRs any antineoplastic therapy within 4 weeks prior to base- are also highly expressed in melanoma cell lines12 and line were excluded from the study. tumour samples obtained from patients with melanoma The study was conducted in accordance with US Food (with some subtypes being more abundant than other and Drug Administration (USFDA) guidelines, Good Clin- 13 subtypes [SSTR1>SSTR2>SSTR3>SSTR4>SSTR5]). The ical Practice, the Declaration of Helsinki and applicable insulin-like growth factor receptors (IGFRs) are also local health authority requirements. All patients provided overexpressed in human melanoma cells, and the IGF-1 written informed consent. The study was approved by the signalling pathway via IGF-1R is shown to play a key role institutional review boards and registered on Clinical - in melanoma progression.14 15 trials. gov as NCT01652547. Pasireotide, a second-generation SSA, has a broader affinity for SSTRs compared with the first-generation study design SSAs, octreotide and lanreotide. The binding affinity of This was an open-label, single-arm, multicentre, intrapa- pasireotide is highest for SSTR5 (IC50: 0.16±0.01 nmol/L) tient, dose-escalation, phase I study to evaluate the prelim- and SSTR2 (IC50: 1±0.1 nmol/L) with moderate affinity inary safety, pharmacokinetics (PK) and antitumour to SSTR3 (IC50: 1.5±0.3 nmol/L) and SSTR1 (IC50: activity of pasireotide (SOM230) subcutaneous (SC) at

9.3±0.1 nmol/L), and low affinity towards SSTR4 recep- doses of 300 µg, 600 µg, 900 µg and 1200 µg three times 16 tors (IC50: >100 nmol/L). Pasireotide also inhibits IGF-1 a day in patients with metastatic melanoma or metastatic plasma levels more effectively than octreotide.17 Pasir- MCC. As per the protocol amendment 4, patients were eotide demonstrated antitumour activity in phase 3 clin- allowed to switch to 80 mg long-acting pasireotide intra- ical studies by significantly reducing growth hormone muscular every 28 days (or a lower dose in case of toxicity) (GH) and IGF-1 levels and shrinking tumours in patients for an additional 6 months, during the follow-up phase 18 with acromegaly. Long-acting pasireotide showed a (online supplementary figure 1). It is to be noted that http://esmoopen.bmj.com/ trend towards higher tumour control rate at month 6 this study was not designed to determine the maximum (although not statistically significant) and was associated tolerated dose (MTD). The maximum allowable dose in with a longer progression-free survival than long-acting this study was 1200 µg three times a day (total daily dose octreotide in patients with metastatic neuroendocrine of 3600 µg, which is still below the MTD for the SC formu- tumours (NETs).19 In a recent phase 2 study in patients lation in healthy volunteers 1950 µg twice a day=total daily with lung NETs, pasireotide alone had significant antitu- dose of 3900 µg). mour efficacy of the same magnitude as everolimus, and the combination of both compounds resulted in a poten- assessments tiation of the antitumour effect.20 The primary objective of the study was to assess the safety on 29 August 2018 by guest. Protected copyright. Therefore, SSTRs and the IGF-1 signalling path- profile of pasireotide SC during the 8-week dose-esca- ways, which are upstream of the Ras/MAPK signalling lation phase. The secondary objectives included the pathway,21 may potentially be targeted using pasireotide assessment of safety at study completion, disease control in BRAF-WT and NRAS-WT melanoma. In the present rate (DCR; defined as the proportion of patients with a study, we assessed the safety profile and the preliminary best overall response (ORR) of complete response (CR) antitumo activity of pasireotide in patients with BRAF-WT or partial response (PR) or stable disease), PK and the and NRAS-WT metastatic melanoma. effect of pasireotide on melanoma response biomarkers (S100b, melanoma-inhibitory activity (MIA)) and treatment response and/or secretion biomarkers over time. PatIents and metHods Safety was monitored by clinical assessments including Adult patients (aged ≥18 years) with histologically/ laboratory evaluations, ECGs, vitals, physical examina- cytologically confirmed, unresectable (stage III) and/ tion, gall bladder ultrasound, and recording adverse or metastatic (stage IV) melanoma or metastatic Merkel events (AEs) through the electronic case report form cell carcinoma (MCC) excluding patients with BRAF and at each visit. The site investigator used RECIST version NRAS mutations were eligible. Only patients with meas- 1.0,22 assisted by the CT or MRI, to assess ORR, defined urable disease according to Response Evaluation Criteria as the proportion of patients with a best overall CR or PR,

2 Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from measured on days 57, 113, 169 and at the end of treat- Table 1 Baseline characteristics ment. In addition, ORR was also assessed using adapted European Organization for Research and Treatment of All patients, Baseline characteristic n=10 Cancer criteria by 18-FDG-PET on days 29 and 57. The serum samples were estimated for PK parameters on days Age (years) 1 and 8 during the four cycles of the dose-escalation phase Median (range) 71.5 (60–77) and at monthly visits during the follow-up phase using Gender, n (%) non-compartmental analysis by Phoenix (V.6.3, Pharsight, Male 8 (80.0) Mountain View, California, USA). Pharmacokinetic parameters included pasireotide maximum/minimum Female 2 (20.0) Race, n (%) concentration on days 1 and 8 for each cycle (Cmax and

Cmin) and area under the curve from 0 to 2 hours (AUC0– Caucasian 10 (100.0)

2h) at steady state. The secretion biomarkers assessed over Ethnicity, n (%) time included S100b and MIA, IGF-1, IGF-2, IGF-binding Other 10 (100.0) protein 2 (IGFBP2) and IGFBP3. Baseline weight (kg) statistical analysis Mean (SD) 77.0 (13.4) A sample size of 18 eligible patients was considered to Height (cm) be sufficient to assess the safety of pasireotide SC in this Mean (SD) 170.6 (8.7) patient population. With 18 patients, there was an 85% Prior antineoplastic therapy, n (%) of probability of detecting an AE with a 10% incidence rate. However, due to the slow recruitment and change in Prior surgery 10 (100) the treatment landscape, this study was terminated early Prior radiotherapy 4 (40) without full enrolment of 18 patients according to the Prior anticancer medications 6 (60) planned sample size. Safety, PK and efficacy results were presented for the 10 enrolled patients. Patients who received at least 1 dose of study medica- those five patients (30%) completed the follow-up phase. tion and had at least one post-baseline safety assessment One patient completed the study as per protocol: 8 weeks were used for all safety analyses. Patients who received at of dose-escalation phase and an additional 6-month least 1 dose of study drug were used for all efficacy anal- follow-up. Two of those three patients (20%) completed yses. The study endpoints were reported according to the follow-up phase receiving 80 mg long-acting pasir- the following two study phases: the dose-escalation phase eotide for at least 3 months. Three of the five patients and the overall phase. The patient characteristics, safety, http://esmoopen.bmj.com/ received pasireotide SC three times a day. efficacy and laboratory endpoints were summarised with descriptive statistics and 95% CI as appropriate. The data were analysed using SASV.9.3. safety Dose-escalation phase The median duration of pasireotide exposure in this Results phase was 6.7 weeks (range, 1.6–8.1 weeks); one patient Patient characteristics was exposed to pasireotide SC for >8 weeks in the Ten patients with BRAF-WT and NRAS-WT melanoma dose-escalation phase. The most commonly reported received the pasireotide treatment. Patients with MCC AEs (incidence ≥20%) were diarrhoea (50%), nausea on 29 August 2018 by guest. Protected copyright. could not be recruited. All the patients were Caucasians (50%), fatigue (20%), hyperglycaemia (20%), hypophos- with the median age of 71.5 years (range, 60–77 years); phatemia (20%), tumour pain (20%) and chills (20%) 80% of the patients were males. All the patients (unre- (table 2). The most frequently reported study drug-re- sectable stage III/IV melanoma) were heavily pretreated, lated AEs (incidence ≥20%) were diarrhoea (50%), underwent prior surgical treatment for their disease nausea (50%), hyperglycaemia (20%) and fatigue and received prior antineoplastic therapies as shown in (20%). Grade 3 or 4 study drug-related AEs (diarrhoea table 1. and nausea) were observed in one patient. Study drug-re- The patient disposition is illustrated in online supple- lated serious AEs (SAEs) were nausea, vomiting and diar- mentary figure 2. The safety, efficacy and PK anal- rhoea in one patient, which occurred at pasireotide SC yses included those 10 patients who were treated with 300 µg. pasireotide. One patient died due to disease progression 5 days after Among the 10 patients, five completed the dose-es- the premature termination of study treatment, and one calation phase; the follow-up phase was reduced from patient died due to disease progression 12 days after the 6 months to 3 months after the decision to terminate withdrawal of consent. Both patients were treated only for the study early, and the patients were offered to receive a very short period of time (16 days and 11 days of treat- further pasireotide treatment in a rollover study. Three of ment, respectively).

Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 3 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from

Table 2 Adverse events (≥20% incidence), regardless of to be study drug related, included nausea, vomiting and study drug relationship, by preferred term and dose level*− diarrhoea in one patient. dose-escalation phase (safety set) efficacy Pasireotide SC, n=10 Among the total 10 patients included in the study, the 300 µg 600 µg 900 µg 1200 µg All best ORR (defined as PR or better) was observed in one tid tid tid tid, patients Preferred term n (%) n (%) n (%) n (%) n (%) (10%) patient (95% CI 0.3 to 44.5), and best overall DCR (defined as stable disease or better) was observed Diarrhoea 4 (40) 1 (10) 0 0 5 (50) in two (20%) patients (95% CI 2.5 to 55.6), both by CT Nausea 3 (30) 2 (20) 1 (10) 0 5 (50) or MRI and 18-FDG-PET. Stable disease and PR were Fatigue 2 (20) 0 0 0 2 (20) observed in one patient each. The best ORR for patients Hyperglycaemia 2 (20) 0 0 0 2 (20) with the measurable disease at baseline is presented in Hypophosphatemia 1 (10) 1 (10) 1 (10) 0 2 (20) (online supplementary table 1). Four patients reached Tumour pain 2 (20) 0 0 0 2 (20) high dosing (two patients reached 3600 µg in the Chills 0 1 (10) 1 (10) 0 2 (20) dose-escalation phase and the follow-up phase, and two patients reached 3600 µg in the dose-escalation phase *List of all AEs that started at each specific dose level. A patient and long-acting pasireotide 80 mg in the follow-up phase with multiple occurrence of an AE preferred term in each dose level and were stable for more than 5 months). One patient is counted only once for that preferred term. AEs, adverse events; SC, subcutaneous; tid, three times a day. achieved stable disease in target lesions (patient number 3). However, it was not considered to be stable disease Overall safety by RECIST criteria due to the worsening of one of the non-target lesions, and the response was considered as Overall, the median duration of exposure to pasireotide progressive disease (PD). SC was 7.6 weeks (range, 1.6–32.1 weeks); one patient was exposed to pasireotide SC for >28 weeks. Of the five Pharmacokinetics patients in the follow-up phase, one patient received The mean plasma concentration of pasireotide versus pasireotide SC and completed the follow-up phase, time profiles on day 8 suggested that the AUC0–2h, Cmax two discontinued the pasireotide SC treatment, and and C increased with the increasing dose. The accumu- two received long-acting pasireotide 80 mg for at least min lation ratio of AUC0–2h was 147% for 300 µg three times 3 months in the follow-up phase. The most commonly a day, and 1%–44% for 600 µg three times a day, 900 µg ≥ reported AEs (incidence 20%) were diarrhoea (50%), three times a day and 1200 µg three times a day, respec- nausea (50%), fatigue (30%), hyperglycaemia (30%), tively. hypophosphatemia (30%), chills (20%), tumour pain

(20%) and decreased weight (20%) (table 3). Study Biomarkers http://esmoopen.bmj.com/ drug-related AEs were reported in 80% of patients, and In patients with observed responses (defined as CR, PR or they included diarrhoea (50%), nausea (50%), hyper- stable disease), the serum levels of S100B and MIA were glycaemia (30%), fatigue (20%), hypophosphatemia sustainably low over the course of the study, whereas in (20%) and weight decrease (20%). The hyperglycaemia some of the patients who had PD, there was a consider- occurring in two patients was managed by concomitant able increase in S100B and MIA serum levels with time. medications including gliclazide and metformin hydro- During treatment with pasireotide, there was a decrease chloride. Grade 3 or 4 drug-related AEs were observed in the serum level of growth factors (IGF-1 and IGF-2) in 20% of patients (diarrhoea and nausea in one patient, and melanoma response biomarkers (MIA and S100B) and hypophosphatemia in one patient). SAEs, suspected (figure 1). Furthermore, there was a decrease in IGFBP3 on 29 August 2018 by guest. Protected copyright. levels (online supplementary figure 3) and an increase in IGFBP2 serum levels with time. Table 3 Adverse events (≥20% incidence) regardless of Ki-67, a marker of proliferation, was assessed before study drug relationship by preferred term (safety set) − overall (dose-escalation phase+follow-up phase) and during pasireotide treatment in four patients. Three of the four tumour samples showed decreases in Ki-67 Preferred term All patients, n (%) staining following pasireotide treatment (figure 2). Diarrhoea 5 (50) Nausea 5 (50) Individual patient response summary The patient characteristics and response to treatment of Fatigue 3 (30) those five patients who completed the treatment in the Hyperglycaemia 3 (30) dose-escalation phase will be described in more detail, Hypophosphatemia 3 (30) as they provide the most relevant information of the Chills 2 (20) potential effect of pasireotide on patients with advanced Tumour pain 2 (20) melanoma. Of these five patients who completed the dose-escalation phase, four patients (patient numbers Weight decreased 2 (20) 3, 4, 8 and 9) received high dose of 3600 µg pasireotide,

4 Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from http://esmoopen.bmj.com/ on 29 August 2018 by guest. Protected copyright.

Figure 1 Individual line plots for biomarkers (A) IGF-1, (B) IGF-2, (C) S100B and (D) MIA. *For patient number 2, samples were collected only for these biomarkers on predose day 12, and on day 1, there were no ‘on treatment’ samples for this patient. CR, complete response; FAS, full analysis set; MIA, melanoma-inhibitory activity; PR, partial response; UNK, unknown. while the highest dose received by the remaining one Patient number 1 was a 64-year-old Caucasian male patient (patient number 1) was lower with 1800 µg. The diagnosed with cutaneous acral-lentiginous melanoma treatment time for the remaining patients may not have (without BRAF and NRAS mutation; stage IVa). The been long enough and/or the applied dose may not have presence of three target and three non-target lesions been high enough in this dose-escalation study to result were identified in this patient. This patient did not in significant responses in efficacy. receive pasireotide treatment as per the dosing schedule

Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 5 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from

Figure 2 Effect on Ki-67 biomarker levels before and after pasireotide. For those patients with the histological data of Ki-67 before and after treatment, n=4. Individual patient response summary; the responses (PR/stable disease/PD) are based on CT/ MRI at specific time points, corresponding to the grey bar. PD, progressive disease; PR, partial response.

mentioned in the study protocol as he did not receive left upper thigh occurred, which was considered as PD. http://esmoopen.bmj.com/ more than 1800 µg and completed four cycles of treat- During cycle 4, the CT/MRI and FDG/PET scan revealed ment. The patient received the first dose of pasireotide worsening of the non-target lesion (pulmonal noduli), 300 µg in the dose-escalation phase. The dose was gradu- and the response was considered as PD. The investigator ally escalated up to 1800 µg as per protocol; the response decided not to discontinue the treatment, as the patient observed was stable disease in the dose escalation phase. was benefitting from treatment. Further evaluations until However, due to the occurrence of grade 3 diarrhoea and the end of study revealed stable disease. This patient also nausea, study drug was temporarily interrupted for 7 days showed significantly lower levels of S100B. At the end of and then reduced to 900 µg. The patient completed the the study, the patient was treated with compassionate use dose escalation phase and was treated for another 12 days of pasireotide without any safety follow-up. on 29 August 2018 by guest. Protected copyright. in the follow-up phase. The overall lesion response was Patient number 4 was a 77-year-old Caucasian male considered as PD in the follow-up phase, which led to the diagnosed with BRAF-WT and NRAS-WT stage IVc cuta- permanent discontinuation of this patient from the study. neous melanoma. The presence of five target lesions and Patient number 3 was a 63-year-old Caucasian female two non-target lesions were identified in the patient. The diagnosed with BRAF-WT and NRAS-WT stage IVa cuta- patient received pasireotide as per the dosing schedule neous nodular melanoma. The presence of one target mentioned in the study protocol and completed six lesion and three non-target lesions were identified in the cycles of treatment. Due to the lack of efficacy, this patient. This patient received pasireotide as per the dosing patient switched to the twice-daily regimen and received schedule mentioned in the study protocol and completed a dose of 2400 µg pasireotide but then discontinued due 10 cycles of treatment with an exception of a dosing error, to disease progression. CT/MRI scans and 18-FDG-PET occurred during the eighth cycle. The results of CT/MRI results showed stable disease for target and non-target scans (on days 29, 57, 113, 169 and 225) and 18-FDG- lesions during cycle 3 (day 30) and PD during cycle 5 PET (on days 29, 57, 113 and 169) revealed the presence (day 58), due to worsening of the non-target lesions. On of stable target lesions throughout the study. However, day 114, CT scan showed worsening of the target lesions, during cycle 3, worsening of the non-target lesion in the continued worsening of the non-target lesion (left deep lung and the generation of a cutaneous lesion in the lymph node) and two new lesions (at left adrenals and

6 Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from liver), and the tumour response was assessed as PD. No dose-escalation phase and two in the treatment follow-up FDG-PET was performed. This patient showed signif- phase). icant improvement in ascites, which appeared to be MAPK and PI3K pathways are the major signalling path- an independent effect of the drug that considerably ways in melanoma, which regulate cell growth, prolifera- improved his quality of life. However, due to disease tion, survival and transformation. These pathways may be progression, he was discontinued from pasireotide treat- activated by the IGF-1 signalling via IGF-1R. Additionally, ment and prescribed with paclitaxel in a therapeutic MAPK pathway may be activated in melanoma directly setting (after 7 days of the last study drug dose). After via activating mutations in the signalling components discontinuation, the ascites recurred, and pasireotide including NRAS, BRAF, MEK 1/2 and c-KIT mutations was readministered that led to a significant reduction in or via mutations in the tumour suppressor genes like weight (from 85 kg to 63 kg) and improvement in ascites NF-1. IGFRs are overexpressed in human melanoma cells, (from grade 2 to grade 1). This patient also showed and the IGF-1 signalling pathway via IGF-1R is shown to significantly increased serum levels of S100B and MIA play a key role in melanoma progression.14 15 Due to the with respect to time. well-established inhibitory effect of pasireotide on GH/ Patient number 8 was a 72-year-old Caucasian male diag- and IGF-1, which was also observed for patients with mela- nosed with BRAF-WT and NRAS-WT stage IVa cutaneous noma in the current study, there may be an opportunity melanoma and had six target lesions and no non-target for combining pasireotide with other anticancer agents lesions. This patient received pasireotide treatment as per targeting different oncogenic pathways in melanoma. KIT the dosing schedule mentioned in the study protocol and inhibitors have shown some efficacy in a subset of patients completed eight cycles of treatment. During all the eval- with BRAF-WT and NRAS-WT melanoma harbouring KIT uations from cycles 1 to 8, the tumour response for this mutations, particularly those with exon 11 or 13 muta- patient was considered as stable disease by CT/MRI (on tions.26 However, this therapeutic option will be appli- days 29, 57, 85 and 114) and FDG/PET (on days 29 and cable only to the 10%–22% of patients with KIT mutant 57), and no new lesions were observed. The patient also BRAF-WT and NRAS-WT melanoma.4 5 showed low S100B and MIA serum levels throughout the In the present study, the antisecretory effect of pasi- study. This patient completed treatment and was further reotide was confirmed by decreased serum levels of enrolled on the rollover study CSOM230B2412. IGF-1 and IGF-2 and a corresponding decrease in Patient number 9 was a 67-year-old Caucasian male IGFBP3 levels (independent of response). Pasireotide diagnosed with BRAF-WT and NRAS-WT stage IV mela- is believed to decrease IGF-1 levels primarily through its noma, identified with the presence of one target lesion ability to decrease GH secretion in the pituitary gland, and one non-target lesion. The patient received pasir- thereby decreasing the synthesis of IGF-1 in the liver.27 eotide as per the dosing schedule of the study protocol The observed suppression of IGF-2 levels suggests that

and completed eight cycles of treatment. No new target pasireotide regulates autocrine/paracrine growth signals http://esmoopen.bmj.com/ lesions were observed during the tumour assessment. in the tumour, since IGF-2 is also known to be secreted CT/MRI (on days 57 and 114) and FDG/PET scans (on by tumours and surrounding cells.28 IGFBP3 is the major days 29 and 60) assessed the tumour response as PR. This carrier of the IGF ligands and is also responsive to GH. patient also showed low levels of S100B and MIA serum The decrease in IGFBP3 provides additional evidence that levels during the study. The patient completed treatment pasireotide had an antisecretory effect. IGFBP2 serum and enrolled in the rollover study CSOM230B2412. level was increased after the treatment with pasireotide, which is supported by similar results observed in patients with acromegaly.17 Additionally, there was a decrease in dIsCussIon the serum level of melanoma response biomarkers (S100B on 29 August 2018 by guest. Protected copyright. This is the first study to demonstrate that pasireotide is and MIA) in patients who had response (defined as PR, safe and well tolerated in patients with BRAF-WT and CR and stable disease), whereas patients with disease NRAS-WT melanoma. The safety profile of pasireotide progression had high S100B and MIA serum levels. These was consistent to that previously reported in other indica- results are supported by the observation that disease tions with the exception of a lower incidence of hypergly- progression is directly proportional to the elevated S100B cemia, observed in this patient population.23–25 A PR and and MIA.29 30 stable disease that were observed in this study comprising Possible limitations of this study include the following: a small sample size (n=10) suggest further investigation insufficient study drug exposure as patients were drop- should be conducted. Of note, patients enrolled in this ping out during the dose-escalation phase due to disease study had significant disease burden (stage IIIC and IV progression. Some patients received a lower dose for a melanoma) and seven patients discontinued treatment shorter period of time and disease progression occurred prematurely (five patients discontinued in the dose-esca- early. Nevertheless, the patients who continued to the lation phase and two in the treatment follow-up phase). follow-up phase received high doses of pasireotide for Most of those who discontinued earlier were not treated more than 5 months and showed stable disease/PR for long and did not reach high doses. Six of them or stable disease in target lesions. Further prospective discontinued due to disease progression (four PDs in the randomised studies are warranted to evaluate the efficacy

Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388 7 Open access ESMO Open: first published as 10.1136/esmoopen-2018-000388 on 23 July 2018. Downloaded from of pasireotide in BRAF-WT and NRAS-WT melanoma. 10. Schally AV, Szepeshazi K, Nagy A, et al. New approaches to therapy of cancers of the stomach, colon and pancreas based on peptide Pasireotide may also be potentially used in combination analogs. Cell Mol Life Sci 2004;61:1042–68. with other therapeutic agents in the treatment of patients 11. Sidéris L, Dubé P, Rinke A. Antitumor effects of somatostatin analogs with BRAF-WT and NRAS-WT melanoma. in neuroendocrine tumors. Oncologist 2012;17:747–55. 12. Martinez-Alonso M, Llecha N, Mayorga ME, et al. Expression of somatostatin receptors in human melanoma cell lines: effect of two Contributors We would like to thank the patients, their families and caregivers and different somatostatin analogues, octreotide and SOM230, on cell the investigators for their participation in this study. We also thank Charu Pundir proliferation. J Int Med Res 2009;37:1813–22. and Rajasree Solipuram (Novartis Healthcare Pvt Ltd) for providing medical editorial 13. Lum SS, Fletcher WS, O'Dorisio MS, et al. Distribution and functional assistance for this research article. significance of somatostatin receptors in malignant melanoma.World J Surg 2001;25:407–12. Funding This study was supported by Novartis Pharmaceuticals Corporation. 14. Lee JT, Brafford P, Herlyn M. Unraveling the mysteries of IGF-1 signaling in melanoma. J Invest Dermatol 2008;128:1358–60. Competing interests RD has intermittent, project-focused consulting and/or 15. Molhoek KR, Shada AL, Smolkin M, et al. Comprehensive analysis advisory relationships with Novartis, Merck Sharp & Dhome (MSD), Bristol-Myers of receptor tyrosine kinase activation in human melanomas reveals Squibb (BMS), Roche, Amgen, Takeda and Pierre Fabre outside the submitted work. autocrine signaling through IGF-1R. Melanoma Res 2011;21:274–84. OM has occasional, project-focused consulting and/or advisory relationships with 16. Bruns C, Lewis I, Briner U, et al. SOM230: a novel somatostatin Novartis, Merck Sharp & Dhome (MSD), Bristol-Myers Squibb (BMS), Roche, Amgen, peptidomimetic with broad somatotropin release inhibiting factor outside of the scope of the submitted work. SG reports grants and other from (SRIF) receptor binding and a unique antisecretory profile.Eur J Novartis, during the conduct of the study; personal fees and other from Intermittent Endocrinol 2002;146:707–16. advisory board relationships with Novartis, Roche, MSD and BMS, outside the 17. Schmid HA, Brue T, Colao A, et al. Effect of pasireotide on glucose- and growth hormone-related biomarkers in patients with submitted work. DS reports grants, personal fees and other from Novartis, during inadequately controlled acromegaly. Endocrine 2016;53:210–9. the conduct of the study; personal fees from Amgen, GSK, BMS, Roche, Amgen, 18. Novartis Pharmaceuticals Corporation. Signifor LAR 2014 Merck, Astra Zeneca, Merck-Serono and Pfizer, outside the submitted work. FC [Prescribing Information]. 2014. https://www.pharma. us. novartis. reports other from Novartis Pharmaceuticals Corporation, outside the submitted com/sites/ www. pharma. us. novartis. com/ files/ signifor_ lar. pdf work. UK-K reports personal fees from Novartis Pharma AG, outside the submitted (accessed 14 Nov 2015). work. HS reports other from Novartis Pharma AG, outside the submitted work. AP 19. Wolin EM, Jarzab B, Eriksson B, et al. Phase III study of pasireotide reports other from Novartis Pharma AG, outside the submitted work. long-acting release in patients with metastatic neuroendocrine tumors and carcinoid symptoms refractory to available somatostatin Patient consent Not required. analogues. Drug Des Devel Ther 2015;9:5075–86. 20. Ferolla P, Brizzi MP, Meyer T, et al. Efficacy and safety of long- ethics approval EK ZH 2012-0209. acting pasireotide or everolimus alone or in combination in patients Provenance and peer review Not commissioned; externally peer reviewed. with advanced carcinoids of the lung and thymus (LUNA): an open-label, multicentre, randomised, phase 2 trial. Lancet Oncol open access This is an open access article distributed in accordance with the 2017;18:1652–64. Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which 21. Iams WT, Lovly CM. Molecular pathways: clinical applications and permits others to distribute, remix, adapt, build upon this work non-commercially, future direction of insulin-like growth factor-1 receptor pathway and license their derivative works on different terms, provided the original work is blockade. Clin Cancer Res 2015;21:4270–7. properly cited, appropriate credit is given, any changes made indicated, and the use 22. Nishino M, Jackman DM, Hatabu H, et al. New Response Evaluation is non-commercial. See: http://creativecommons. org/ licenses/ by- nc/ 4. 0/. Criteria in Solid Tumors (RECIST) guidelines for advanced non- small cell lung cancer: comparison with original RECIST and impact on assessment of tumor response to targeted therapy. AJR Am J Roentgenol 2010;195:W221–W228. 23. Silverstein JM. Hyperglycemia induced by pasireotide in patients with Cushing's disease or acromegaly. Pituitary 2016;19:536–43. http://esmoopen.bmj.com/ REFERENCES 24. Cives M, Kunz PL, Morse B, et al. Phase II clinical trial of pasireotide 1. Ekedahl H, Cirenajwis H, Harbst K, et al. The clinical significance of long-acting repeatable in patients with metastatic neuroendocrine BRAF and NRAS mutations in a clinic-based metastatic melanoma tumors. Endocr Relat Cancer 2015;22:1–9. cohort. Br J Dermatol 2013;169:1049–55. 25. Gadelha MR, Bronstein MD, Brue T, et al. Pasireotide versus 2. Jakob JA, Bassett RL, Ng CS, Cs N, et al. NRAS mutation status is continued treatment with octreotide or lanreotide in patients with an independent prognostic factor in metastatic melanoma. Cancer inadequately controlled acromegaly (PAOLA): a randomised, phase 3 2012;118:4014–23. trial. Lancet Diabetes Endocrinol 2014;2:875–84. 3. Cancer Genome Atlas Network. Genomic Classification of Cutaneous 26. Guo J, Si L, Kong Y, et al. Phase II, open-label, single-arm trial of Melanoma. Cell 2015;161:1681–96. imatinib mesylate in patients with metastatic melanoma harboring 4. Hodis E, Watson IR, Kryukov GV, et al. A landscape of driver c-kit mutation or amplification. Journal of Clinical Oncology mutations in melanoma. Cell 2012;150:251–63. 2011;29:2904–9. 5. Mar VJ, Wong SQ, Li J, et al. BRAF/NRAS wild-type melanomas 27. Schmid HA. Pasireotide (SOM230): development, mechanism on 29 August 2018 by guest. Protected copyright. have a high mutation load correlating with histologic and molecular of action and potential applications. Mol Cell Endocrinol signatures of UV damage. Clin Cancer Res 2013;19:4589–98. 2008;286:69–74. 6. Pawlikowski M, Melen-Mucha G. Perspectives of new potential 28. Li B, Tsao SW, Chan KW, et al. Id1-induced IGF-II and its autocrine/ therapeutic applications of somatostatin analogs. Neuro Endocrinol endocrine promotion of esophageal cancer progression and Lett 2003;24:21–7. chemoresistance--implications for IGF-II and IGF-IR-targeted 7. Heaney AP, Melmed S. Molecular targets in pituitary tumours. Nat therapy. Clin Cancer Res 2014;20:2651–62. Rev Cancer 2004;4:285–95. 29. Domingo-Domènech J, Molina R, Castel T, et al. Serum protein s-100 8. González-Barcena D, Schally AV, Vadillo-Buenfil M, et al. Response predicts clinical outcome in patients with melanoma treated with of patients with advanced prostatic cancer to administration of adjuvant interferon--comparison with tyrosinase rt-PCR. Oncology somatostatin analog RC-160 (vapreotide) at the time of relapse. 2005;68:341–9. Prostate 2003;56:183–91. 30. El Fitori J, Kleeff J, Giese NA, et al. Melanoma Inhibitory Activity 9. Froidevaux S, Eberle AN. Somatostatin analogs and radiopeptides in (MIA) increases the invasiveness of pancreatic cancer cells. Cancer cancer therapy. Biopolymers 2002;66:161–83. Cell Int 2005;5:3.

8 Dummer R, et al. ESMO Open 2018;3:e000388. doi:10.1136/esmoopen-2018-000388